The economics of human trafficking and labour ... - WordPress.com

Viewer
Transcript

Journal of Comparative Economics 38 (2010) 173–188

Contents lists available at ScienceDirect

Journal of Comparative Economics journal homepage: www.elsevier.com/locate/jce

The economics of human trafﬁcking and labour migration: Micro-evidence from Eastern Europe Toman Omar Mahmoud a,b,*, Christoph Trebesch c,d a

Kiel Institute for the World Economy, Hindenburgufer 66, 24105 Kiel, Germany German Institute of Global and Area Studies, Neuer Jungfernstieg 21, 20354 Hamburg, Germany c Free University of Berlin, Binger Str. 40, 14197 Berlin, Germany d Hertie School of Governance, Friedrichstraße 180-184, 10117 Berlin, Germany b

a r t i c l e

i n f o

Article history: Received 12 June 2009 Revised 20 January 2010 Available online 13 February 2010 JEL classiﬁcation: F22 J61 K42 O17 Keywords: Human trafﬁcking Migrant exploitation Illegal migration Migration networks Eastern Europe

a b s t r a c t Omar Mahmoud, Toman, and Trebesch, Christoph—The economics of human trafﬁcking and labour migration: Micro-evidence from Eastern Europe Human trafﬁcking is a humanitarian problem of global scale, but quantitative research on the issue barely exists. This paper is the ﬁrst attempt to analyze the economics of human trafﬁcking and labour migration based on micro data, using unique household surveys from Belarus, Bulgaria, Moldova, Romania, and Ukraine. We ﬁnd that individual trafﬁcking risks are much higher in regions with large emigration ﬂows. The reasons are lower recruitment costs for trafﬁckers in emigration areas and, to a less extent, more negative self-selection into migration. Our results also indicate that illegal migration increases trafﬁcking risks and that better information, e.g. through awareness campaigns, might be an effective strategy to reduce the crime. These ﬁndings may help policymakers to better target antitrafﬁcking efforts. Journal of Comparative Economics 38 (2) (2010) 173–188. Kiel Institute for the World Economy, German Institute of Global and Area Studies, Germany; Free University of Berlin, Hertie School of Governance, Germany. Ó 2010 Association for Comparative Economic Studies Published by Elsevier Inc. All rights reserved.

‘‘Like many things that should have been stamped out a long time ago, slavery, it seems, is alive and well.” The Economist, 9 March 2005 ‘‘The trade in people is surely the most morally repugnant of all the illicit trades that ﬂourish today. But it’s deeply entrenched and interwoven with the world’s ever more complex migration ﬂows.” Moises Naim in ‘‘Illicit” (2005, p. 89). 1. Introduction The exploitation and trafﬁcking of men, women and children is a humanitarian problem of global scale. A recent study estimates that at least 12 million people worldwide are trapped in conditions of forced labour. Around a ﬁfth of these are being exploited as a result of human trafﬁcking (ILO, 2005). These forms of modern day slavery have become one of the most proﬁtable businesses in the world – and one of the most horrifying. Human trafﬁcking is said to be the fastest growing source of income for organized crime and its third most important, exceeded only by drugs and arms trade (Obuah, 2006). * Corresponding author. E-mail addresses: [email protected] (T. Omar Mahmoud), [email protected] (C. Trebesch). 0147-5967/$ - see front matter Ó 2010 Association for Comparative Economic Studies Published by Elsevier Inc. All rights reserved. doi:10.1016/j.jce.2010.02.001

174

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188

After decades of neglect, public attention on the problem has increased considerably in recent years. Policymakers in international organisations and national governments all over the world as well as NGOs have augmented their efforts to combat trafﬁcking and to raise public awareness on the issue.1 There has also been a notable increase in donor funding, e.g. by the European Union and the US government. Despite the policy relevance, academic research on the topic remains scarce, especially when it comes to economics. The large economic literature on organized crime and illicit activities (e.g. Fiorentini and Peltzman, 1995 or Freeman, 1999) has largely ignored the sizeable market of human trafﬁcking and migrant exploitation. The same is true for the rapidly growing ﬁeld on the economics of migration, which has been notably silent on the issue. As a result, and despite frequent calls for empirical research on the topic,2 there is still very little knowledge on human trafﬁcking as an economic phenomenon. One major reason for this is the grave difﬁculty in gathering reliable and representative data. This paper is the ﬁrst to analyze the drivers of human trafﬁcking with micro data. Our theoretical argument is based on simple supply and demand considerations and rooted in the literature on the economics of crime and migration. In a nutshell, we argue that trafﬁcking and exploitation are the sad but obvious consequence of migration pressure in a world of closed borders. The growing, but unsatisﬁed demand for legal migration options has created a breeding ground for criminal organisations and exploitative employers, who have learned to make a proﬁt from people’s desire to work abroad. Migration has become a multibillion-dollar industry that can be particularly lucrative for those willing to use physical violence and restraint. Our focus on the close link between trafﬁcking and labour migration differs from that of many recent media and policy reports.3 Most reports point to legislation and law enforcement, poor border controls, bribery and corruption, or insufﬁcient education as drivers of human trafﬁcking. These factors certainly exacerbate the trafﬁcking problem, but may not explain it at its core. We argue instead that it is ﬁrst and foremost the wish for a better life abroad that puts millions of people at the risk of ending up in exploitative work conditions. Their willingness to depart and to take risks in the migration process can be easily exploited by criminal agents. In line with the above, we expect the risk of human trafﬁcking to be highest in areas with high rates of emigration. Simply put, ‘‘trafﬁckers ﬁsh in the stream of migration” (Coomaraswamy, 2001, p. 3), so that more people leaving means more people at risk. In addition, two channels may explain why individual trafﬁcking risks increase with regional emigration rates. The ﬁrst channel relates to the demand side of the human trafﬁcking market (criminal agents recruiting for employers abroad). We predict that recruitment costs for trafﬁckers will be lower in areas of large-scale emigration. One reason is that it is easy to ﬁnd potential victims, simply because so many people leave from there. The other reason is the agglomeration of shadow migration industries. Large migrant ﬂows attract a variety of middlemen offering services such as false documents, smuggling across borders or work procurement abroad. The presence of these agents offers ample opportunities for trafﬁckers as they can collaborate or free-ride on their reputation (Tamura, 2007). The second channel is linked to the supply side of the human trafﬁcking market (vulnerable individuals willing to work abroad). Migrants departing from high-migration areas are likely to be negatively self-selected and share characteristics that make them more vulnerable to trafﬁcking. High-emigration areas are often disadvantaged in the ﬁrst place, so that migrants from there may be more willing to take risks, and possibly even consent to hazardous working conditions abroad. It has also been shown that large-scale emigration can lower the costs of migration, thus fostering the migration of poorer, less educated workers (McKenzie and Rapoport, forthcoming; Beine et al., forthcoming). In sum, we predict trafﬁcking risks to be higher in high-migration areas (i) because trafﬁckers beneﬁt from lower recruitment costs and free-riding opportunities there and/or (ii) because those departing from high-migration areas can be more easily deceived. To test our arguments, we use a novel and unique survey on human trafﬁcking which covers 5513 randomly selected households from 82 regions in Belarus, Bulgaria, Moldova, Romania, and Ukraine. These Eastern European countries are among the most important source countries of human trafﬁcking worldwide (ILO, 2005). Human trafﬁcking is deﬁned as a situation in which an individual travelling abroad was locked and forced to work for no or little pay via means of coercion. As such, trafﬁcking differs importantly from migrant smuggling, a commercial service that normally occurs with the consent of migrants, and from illegal migration, which does not typically involve any forms of exploitation. About 7% of the migrant families (108 out of 1563 households) identiﬁed by the survey reported to have a trafﬁcking victim among their members. This includes victims of both sexual and non-sexual forms of human trafﬁcking, but not trafﬁcked children. To supplement our ﬁndings, we present an additional case study of Moldova, which has by far the highest prevalence of human trafﬁcking among the ﬁve countries under study. The case study is based on a second dataset covering 1679 migrants from 211 localities in Moldova. Although this second survey does not explicitly focus on trafﬁcking, it allows identifying worst forms of migrant exploitation and contains ample additional information on migrants’ characteristics including their destination countries.

1 Most prominently, the UN issued the Protocol to Prevent, Suppress and Punish Trafﬁcking in Persons (UN, 2000a) and the Protocol Against the Smuggling of Migrants (UN, 2000b), which entered into force in 2004 and 2003 respectively. In March 2007, the UN also formally launched a Global Initiative to Fight Human Trafﬁcking (UN.GIFT). The work of counter-trafﬁcking NGOs such as Free the Slaves or La Strada and pioneering books such as Bales (1999) have helped to increase awareness of the problem. 2 Gozdziak and Collett (2005, p.122) conclude that ‘‘future research needs to move beyond stating that there is a problem, to more systematic and rigorous data collection and analysis.” Among others, similar calls have been made by Tyldum and Brunovskis (2005), Piper (2005) and Chand (2008). 3 Compare e.g. ILO (2005), UNODC (2006) and US State Department (2008).

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188

175

The key result is in line with our expectations. Individual trafﬁcking risks increase with the size of migration ﬂows from the region (ﬁve-country survey) or locality (Moldova case study). Migration prevalence and other proxies of migration pressure turn out to be key predictors of trafﬁcking on the household level. Thus, we ﬁnd that trafﬁcking tends to occur where migration ﬂows are largest, not in remote regions with no or little migration. This result holds when controlling for selection effects and a broad range of household and migrant attributes as well as characteristics of the locality, region and country. As to the potential channels explaining this ﬁnding, we ﬁnd stronger evidence for demand-side than for supply-side factors. Beyond this, we can report two further main results. First, illegal migration patterns increase trafﬁcking risks. Second, we present evidence that awareness campaigns could play a successful role in reducing trafﬁcking risks. The incidence of trafﬁcking is considerably lower in regions where the awareness of the phenomenon of human trafﬁcking is high. The rest of the paper is structured as follows: Section 2 deﬁnes the main concepts and summarizes the current state of knowledge on human trafﬁcking. It also provides stylized facts on the patterns of migration and trafﬁcking in Eastern Europe. Section 3 sketches our main argument. In Section 4, we describe the data and our empirical strategy. Section 5 discusses the results and presents a number of robustness checks. Section 6 presents the case study on Moldova. Section 7 concludes.

2. What do we know about human trafﬁcking? 2.1. Deﬁnitions It is crucial to clearly deﬁne the concepts that lie at the core of this paper. We follow the UN protocols against human trafﬁcking, which, after years of debate, succeeded in clearly deﬁning and distinguishing these concepts.4 In accordance with the UN deﬁnition, the main purpose of smuggling is to facilitate the illegal entry of a person into another country. Typically, migrants are smuggled with their consent, paying the smuggler for his service. The smuggler gains a material beneﬁt, but does not necessarily exploit the migrant. In contrast, trafﬁcking always involves the violation of human rights and severe forms of exploitation. Victims of trafﬁcking are recruited, transported and forced to work by means of coercion. The main purpose of trafﬁcking is exploitation, either sexual exploitation or other forms of economic exploitation. While these deﬁnitions are helpful from a legal perspective, trafﬁcking, exploitation and smuggling are often interrelated.5 Some migrants using smuggling services may depart voluntarily, but end up being exploited. In such cases, smugglers become trafﬁckers. 2.2. Determinants of human trafﬁcking A fundamental breeding ground for trafﬁcking and exploitation is the economic situation of people in poorer regions of the world, pushing vulnerable people to emigrate and seek better opportunities abroad (Bruckert and Parent, 2002; Chand, 2008). The large potential gains from migration,6 joint with network and herd effects have generated an unprecedented demand for legal migration to richer countries (World Bank, 2005). However, with most middle and high-income country labour markets shut off, there is only very limited supply of legal employment opportunities abroad. At the same time, there is a constant demand for cheap manual workers and prostitutes in both developed and developing countries. This situation has fostered the emergence of shadow migration industries offering services such as border crossings and the procurement of illegal work abroad (Orrenius, 1999; Stalker, 2000; Gathmann, 2008). These activities, including its worst forms such as human trafﬁcking, offer large monetary rewards and have to be seen in parallel to other ﬁelds of criminal intermediation, e.g. drugs and arms trade. Aronowitz (2001), Salt and Stein (1997) and Schloenhardt (1999) provide a detailed analysis of the business of trafﬁcking and exploitation. In essence, the organisational structure of trafﬁcking networks is similar to that of providers of legal services. Trafﬁcking organisations ﬁnd and attract people willing to work abroad via advertisements in newspapers, the internet or through employment agencies. They also contact people on an individual basis, often via dispersed recruiters who may have even been trafﬁcking victims themselves. All available evidence indicates that human trafﬁcking is an exceptionally lucrative business for criminal groups. In a recent study, the ILO (2005) estimates that sexual and labour exploitation yields US$ 32 billion of proﬁts a year to the actors involved. This corresponds to an estimated US$ 13,000 of yearly proﬁts for each forced labourer. Another estimated ﬁgure is the proﬁt of criminal gangs from sex trafﬁcking alone, which is conservatively estimated at US$ 5–7 billion a year, with Interpol giving a higher estimate of US$ 19 billion annually (ILO, 2005; US State Department, 2008). The business can be highly attractive for criminal actors. Start-up costs are small and as Väyrynen (2005) harshly puts it, ‘‘people are a good commodity as they do not easily perish, but they can be transported over long distances and can be reused and re-sold”. Moreover, the risks of detection, prosecution or arrest are much lower compared to other ﬁelds of illegal activity such as drug or arms trade. Exploited victims often feel discouraged to denounce their trafﬁckers, as they face the risk of deportation and other legal consequences when contacting authorities in receiving countries (Langberg, 2005; Sur-

4

The exact deﬁnitions are given in UN (2000a,b). In this paper, we use the terms ‘‘trafﬁcking” and ‘‘exploitation” interchangeably. 6 For an individual, emigration can be the single most effective strategy to improve economic opportunities and increase income, often by a multiple. The large ‘‘place premium” in wages between sending and receiving countries is carefully documented in a recent paper by Clemens et al. (2008). 5

176

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188

tees, 2005). And even if trafﬁckers are arrested, penalties are relatively low compared to other illicit activities (LexisNexis, 2008). In economic research, two recent theoretical papers have shed some light on the link between migration and its dark side in the form of trafﬁcking or forced labour.7 Tamura (2007) focuses on the interaction of human smuggling and trafﬁcking. In his model, migrants hire a smuggler to cross borders and ﬁnd work abroad. Once migrants depart, it depends on the smuggler’s decision and the proﬁtability of exploitation whether they end up being trafﬁcked or not. In a different setup, Friebel and Guriev (2006) model the market of illegal migration with debt/labour contracts. As most migrants cannot pay for migration costs in advance, criminal intermediaries and smugglers offer loans to potential migrants, which they have to pay back in the destination country. The contracts between migrants and intermediaries can only be enforced in the illegal sector, not in the legal one. Although the enforcement of such contracts can take place through coercion and punishment, the possible risk of exploitation and trafﬁcking by the intermediaries is ruled out. While theory is scarce, empirical evidence is even scarcer. There is very little systematic knowledge about which households and regions are most vulnerable to exploitation and human trafﬁcking. We are aware of only one econometric study on human trafﬁcking.8 Akee et al. (2007) analyze the determinants of trafﬁcking of children and women in a crosssection of countries. A main result is that the stage of economic development and inequality of incomes between countries are major determinants of trafﬁcking. Contrary to expectations, they ﬁnd that granting legal status to trafﬁcked victims in host countries and banning prostitution in source countries leads to a higher, not lower, likelihood of trafﬁcking. Despite this pioneering study, much remains to be done to uncover the drivers and mechanisms of human trafﬁcking, particularly on the micro-level. 2.3. Human trafﬁcking and labour migration in Eastern Europe There is a small but growing body of literature on the general migration patterns in Eastern Europe.9 The countries under study have witnessed large and often seasonal migration ﬂows since the breakdown of the Soviet Union. The main drivers of migration in the area are poor living conditions and a lack of jobs as well as migration networks. There is little evidence for positive self-selection, as most migrants work in low-skilled jobs in sectors such as construction, agriculture or in domestic services, often under dismal conditions. As to destinations, Mansoor and Quillin (2006) estimate that from 1990 to 2006 about 80% of emigrants from the former Soviet Union moved within the CIS region, predominantly towards Russia, which is now the second largest recipient of international migrants worldwide. Other main destinations, particular for migrants from Bulgaria and Romania, are countries in the European Union. Countries of the former Eastern bloc have also become major origins of forced labour and human trafﬁcking in the last two decades. The exploitation of human beings has a long and sad history in the region. In medieval times, Venetian and Genoese merchants established a massive trade with slaves from Eastern Europe and the epoch of ‘‘second serfdom” implied highly repressive forms of exploitation (Blum, 1957; Domar, 1970). Today, the main destinations of trafﬁcking victims are Western Europe and Russia, but they also include North America, Asia and increasingly the Middle East, in particular Turkey, Israel, Lebanon and the United Arab Emirates (Surtees, 2005; UNODC, 2006). Estimates of the extent of trafﬁcking and forced labour in Eastern Europe vary widely. The report by the ILO (2005) estimates that a minimum of 200,000 people from Eastern Europe and Central Asia have fallen victim to trafﬁcking, while Mansoor and Quillin (2006) cite estimates from the United Nations Population Fund that more than 175,000 persons are trafﬁcked annually in Europe and Central Asia. Media coverage and reports such as UNODC (2006) suggest that Eastern European trafﬁcking victims are mostly young women who are exploited in the commercial sex industry. It is certain that many victims from the region are females forced into prostitution. Yet, the IOM (2006) report and recent more qualitative surveys from Russia and South-eastern Europe (Tyuryukanova, 2005; Surtees, 2005) underline the dominant but neglected issue of male exploitation.10 According to these sources, a large share of those trafﬁcked from Central and Eastern Europe are men being exploited in agriculture, construction, or warehouse work. A further key insight from recent studies in the region is that most of the victims of trafﬁcking depart on a voluntary basis (ILO, 2005; Tyuryukanova, 2005). They often enter the destination countries at legal border crossings and with legal documents (Surtees, 2005; ILO, 2005). Typically, victims become subject to coercion and violence only after arrival at the destination, where their life in the shadows and a new environment makes them prone to be lured into seemingly attractive jobs by trafﬁckers. The role of middlemen in the Eastern European trafﬁcking business has also been documented in a small number of studies. According to the ILO (2005), trustworthy social networks can lower the risks of being trafﬁcked, while the use of middlemen increases it. However, it is important to underline that social networks do not assure safety from trafﬁcking. The IOM study by Surtees (2005) illustrates that many victims of sexual exploitation in countries like Albania, Bulgaria, or

7 In a further paper, Dessy and Pallage (2006) focus on the trade off between trafﬁcking and child labour. In their model, the risk of trafﬁcking acts as a deterrent to sending children to work. They come to the conclusion that successful policy measures against trafﬁcking may lead to an increase in child labour, because parents will be less afraid of letting their children work. 8 Basu and Chau (2003) provide cross-country evidence on child labour in debt bondage, but do not place their analysis in a migration or trafﬁcking context. 9 The following stylized facts are based on Bauer and Zimmermann (1999), Kraler and Iglicka (2002), Mansoor and Quillin (2006) and Görlich and Trebesch (2008). 10 Piper (2005) highlights that there is a particular need for data and empirical research on non-sexual forms of trafﬁcking.

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188

177

Macedonia were recruited by friends. Similarly, Lackso and Gramegna (2003) report that in South-eastern Europe 60% of victims are recruited by acquaintances, mostly with the promise to ﬁnd them a job abroad. In sum, these stylized facts show that the human trafﬁcking ﬂows from the region run in parallel to the general migration ﬂows in Eastern Europe and trafﬁcking affects men and women alike. It also appears that many trafﬁcked victims leave voluntarily, but ill-prepared.

3. Human trafﬁcking as a side effect of international labour ﬂows This section discusses our theoretical argument and the related channels at work. In line with the above, we argue that human trafﬁcking is an unavoidable side effect of migration in a world with large income disparities but closed borders. We therefore predict that large migrant outﬂows will increase the incidence of trafﬁcking in a region. Simply put, more people departing means more people at risk. Assuming a ﬁxed (low) probability that a departing migrant ends up being trafﬁcked, it is straightforward to expect a larger number of trafﬁcking victims to come from regions with more emigration. Our argument, however, goes beyond this purely mechanical effect. We predict that individual trafﬁcking risks increase with the scale of regional emigration. Two channels can explain this. The ﬁrst channel is linked to the demand side of the human trafﬁcking market, as mass emigration can lower recruitment costs of trafﬁckers. The second channel relates to the supply side of the trafﬁcking market, as those departing from high-emigration regions may share characteristics that make them more prone to be trafﬁcked. On the demand side, large-scale emigration may create agglomeration and free-riding opportunities for trafﬁckers. Just like legal businesses, criminal activities tend to agglomerate in larger markets. This is true for drug gang concentration in cities (Glaeser and Sacerdote, 1999) and applies to many other crime industries that concentrate geographically (Freeman et al., 1996; Zenou, 2003). In the context of trafﬁcking, markets will be larger and more attractive in areas of mass emigration which offer high economies of scale. Most importantly, these regions feature a continuous supply of potential victims, which can considerably ease recruitment. Additionally, trafﬁckers can beneﬁt from the existence of a local shadow migrant industry, in which private agents offer all kinds of work procurement or travel services to emigrants. Related industries have thrived in Northern Africa or at the US–Mexican border, but also in many emigration regions of Eastern Europe. In such settings, trafﬁckers can take advantage of agglomeration effects by collaborating with, or free-riding on, other agents such as people smugglers or travel agencies (see Tamura, 2007).11 The result will be a higher concentration of trafﬁckers and, in turn, higher trafﬁcking risks in such regions. On the supply side, it is mostly negative self-selection that could explain higher risks in emigration areas. Localities witnessing large-scale out-migration may be particularly disadvantaged in the ﬁrst place, often suffering from poor economic development and weak institutions. In deprived areas, migration may be the only strategy available to increase economic opportunities and start a better life. This will have implications for the preference set of those departing, as they may be more prone to take risks in the migration process, such as crossing borders illegally or accepting jobs in the informal sector abroad. In extreme cases, migrants from particularly disadvantaged regions may even consent to exploitative working conditions. They may be keen to access so-called ‘‘3D” jobs (difﬁcult, dirty, dangerous) because they would do nearly everything just to escape their dismal living conditions at home (Bhabha, 2005). Besides such risk-taking behaviour, skills and education may also play a role. The issue of negative or positive self-selection into migration is highly controversial.12 However, recent studies agree in that larger migration ﬂows can foster negative self-selection (Beine et al., forthcoming; Chiquiar and Hanson, 2005; McKenzie and Rapoport, forthcoming). Larger networks and an established migration infrastructure lower the costs of migration, thereby increasing the likelihood that more people from lower social classes start to emigrate. These less-qualiﬁed late-comers in the migration wave may be more vulnerable to deception schemes and less able or willing to thoroughly prepare their journey. On the supply side we therefore predict trafﬁcking risks to be higher in high-migration areas because those departing from there may be more desperate or simply less qualiﬁed to handle the many perils of the migration process. Stretching even further, one could imagine the ﬁrst and second channel to mutually reinforce each other. Negative selfselection may be beneﬁcial for trafﬁckers, and pose an additional incentive to agglomerate in regions of mass emigration and large networks. The reason is that negative self-selection may have relevant feedback effects on network composition or network quality. The more low-skilled workers depart, the worse the migrant network might become with regard to information ﬁltering and transmission and in safeguarding from exploitation. As a result, larger, and thereby less-skilled, regional migrant networks might facilitate recruitment. In conclusion, both supply and demand-side channels predict individual trafﬁcking risks to be higher in areas of largescale emigration. Note, however, that this reasoning stands in contrast to some of the migration literature. Most authors in the ﬁeld suggest that risks should be lower in areas of mass emigration, mainly because of the protective role of migrant networks. Community-level networks are often assumed to decrease information asymmetries and facilitate the arrival at 11 Recent research indicates that the markets of migrant smuggling, work procurement and human trafﬁcking/exploitation are indeed closely intertwined (e.g. Leman and Janssens, 2007; Väyrynen, 2005; Aronowitz, 2001; Ruggiero, 1997). 12 Recent years have seen evidence for positive, intermediate and negative self-selection into migration, depending on the data and country investigated. Chiquiar and Hanson (2005), Mora and Taylor (2005), Taylor et al. (2003), or Adams (2005) ﬁnd years of schooling and other human capital variables to have a signiﬁcant positive impact on the decision to migrate. Orrenius and Zavodny (2005) ﬁnd evidence for intermediate selection in Mexican migration to the US, while Borjas (1987) predicts negative self-selection from countries with high earnings inequality such as Mexico.

178

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188

the destination abroad, thereby reducing risks (see e.g. Carrington et al., 1996; Winters et al., 2001; Munshi, 2003; Bauer et al., 2007). We argue, however, that the role of networks may be ambiguous, especially when external agents learn to misuse them. This view is in line with some recent sociological research showing how exploitative employers learn to take advantage of migrant networks. One example is the study on ‘‘Networks of Exploitation” (Cranford, 2005) which shows that employers in the Los Angeles janitorial industry rewarded Mexican supervisors to recruit vulnerable workers from their home communities into dismal jobs. Migrant networks were systematically misused to replace unionized workforce with immigrants, who would earn much less and could be more easily exploited. Further studies with related arguments and evidence are Martin et al. (2007) who study the exploitation of immigrants’ arriving in the Chicago South Side, and Zhang and Chin (2002) who track the decisive role of networks in Chinese human smuggling. Thus, it seems that networks on the regional or local level do not necessarily shield from the (low) risk of being trafﬁcked. 4. Data and empirical strategy 4.1. Data The main body of analysis uses a unique household survey which was commissioned by the International Organization for Migration and covers ﬁve Eastern European countries. In addition, we also employ a second household survey from Moldova to check the robustness of our results and provide supplementary evidence. This survey is described in Section 6. Fieldwork for the ﬁve-country survey was conducted in Belarus, Bulgaria, Moldova, Romania and Ukraine between August and September 2006. To arrive at nationally representative estimates, the survey relied on a multi-stage stratiﬁed sampling scheme.13 Accordingly, the survey covers both urban and rural areas and sampled households of all regions in the ﬁve countries.14 In each country, at least 1000 individuals were interviewed, with slightly larger samples in Bulgaria and Ukraine, resulting in a total sample of 5513 households. The interviews were conducted with the person whose birthday was last among the rest of dwellers. The largest module in the survey was dedicated to human trafﬁcking across borders. Most importantly, interviewed persons were asked whether persons in their close surroundings (relatives, friends and neighbours), in their close family,15 or they themselves had experienced situations which would be classiﬁed as trafﬁcking according to the deﬁnition above. Since households were randomly selected for the interview, selection bias suffered by other datasets that are based on the number of assisted victims or the number of border apprehensions (Tyldum and Brunovskis, 2005) is avoided here. To the best of our knowledge, the relatively large sample and the survey design make it the most comprehensive and representative micro-level survey on human trafﬁcking available worldwide. 4.2. Dependent variable and econometric model Our dependent variable is a dummy denoting whether the interviewed person reported a victim of human trafﬁcking among close family members.16 Given the relatively small number of individuals with trafﬁcking in their close family, we combine sexual exploitation and other forms of exploitation. More speciﬁcally, the dummy takes the value of one if a close family member had travelled abroad and experienced one of the following situations: (i) ‘‘was offered a domestic or nursing job, but was locked and forced to work for no pay”, (ii) ‘‘was offered a job at an enterprise, on a construction site, or in agriculture, but was locked and forced to work for no or little pay” and (iii) ‘‘was offered employment, but the passport was taken away upon arrival to the destination country and was forced to work in the sex business”. In line with the UN deﬁnition of human trafﬁcking given above, these three situations have in common that the individual was forced to work by means of coercion. Overall, 108 individuals indicate a victim of human trafﬁcking in their close family. Out of these, there are 22 reported cases of members being forced to work in the sex business, while 86 cases were associated with non-sexual forms of exploitation. In our sample, most trafﬁcked cases come from Moldova (56 out of 1073). The rest is found in Bulgaria (18 out of 1007), Ukraine (15 out of 1345), Belarus (11 out of 1071), and Romania (8 out of 1017). A very large number of interviewed people had not heard about any case of trafﬁcking among the people they knew (4905 out of 5513). Our data are likely to suffer from some degree of reporting bias. Victims of human trafﬁcking and their relatives might be reluctant to share their experiences, being afraid of stigmatisation or problems with authorities at home or abroad. This is especially true for the case of sexual exploitation, which is a highly sensitive issue (Laczko, 2005). To minimize the degree of under-reporting, the survey instrument used various ways of reducing the stigma associated with human trafﬁcking. First, instead of asking directly whether a family member had been ‘‘trafﬁcked”, the questionnaire rather referred to typical situations of coerced labour. Second, the respondent did not have to reveal the identity of a trafﬁcked family member. Third, if there was a case of sexual exploitation in the family, the interviewed person could still ‘‘hide” this experience in a more gen13

See IOM (2006) for more details. Overall, the analysis covers 82 regions. The term ‘‘close family” is deﬁned as only including parents, children, husband/wife, brothers and sisters of the interviewed person. The survey generally refers to households as identiﬁed by close family ties with respect to the interviewee and not by residency. 16 As the interviewee is by deﬁnition part of the close family, we make sure that our dependent variable also includes the case when the interviewed person was a victim of human trafﬁcking him- or herself. 14 15

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188

179

eral answer category of exploitation. Based on these considerations, we believe that the potential under-reporting bias should not seriously invalidate our ﬁndings.17 We estimate the determinants of human trafﬁcking on the household level with binary response models. A potential problem in this regard is that human trafﬁcking, although affecting millions of people, is a relatively rare event on the household level. King and Zeng (2001) show that the maximum likelihood estimator of standard logit or probit models is biased when the number of non-events (zeros) is multiple times higher than the number of events (ones). They illustrate that standard binary models underestimate the probability of rare events and do increasingly so as the event gets rarer. As a baseline method, we therefore rely on the rare event logit model suggested by King and Zeng (2001). In essence, their model is based on the basic logit model, but uses a bias-corrected estimator for rare events that generates a lower mean square error for coefﬁcients. Note that this approach is particularly popular in political science (e.g. Fearon and Laitin, 2003; Harff, 2003; Collier and Hoefﬂer, 2004) but is also increasingly used in research on economic issues (e.g. Hausmann et al., 2005; Leblang and Satyanath, 2008). To test the robustness of our results, standard logit models are also applied.18 4.3. Identifying the channels at work Our empirical approach follows the theoretical arguments above. We ﬁrst test the general mechanical prediction that trafﬁcking is more likely to occur in regions of large-scale emigration. We then focus on the subsample of migrant households to establish whether the individual likelihood of being trafﬁcked increases with regional migration prevalence. In a ﬁnal step, we investigate the channels behind our ﬁnding on the link between trafﬁcking and migration ﬂows. More speciﬁcally, we use proxies of regional recruitment conditions (demand-side factors) and proxies indicating the skill level or desperation of those departing (supply-side factors) and interact them with regional emigration rates. This will provide some indication on whether mass emigration reinforces the demand and supply-side channels described above. Our key variable of interest is the regional migration prevalence rate. It is constructed by taking the share of surveyed households in each region that reported to have had a member of the close family abroad in the last 3 years. As an alternative, we employ a proxy of low migration pressure. Concretely, we construct a regional variable capturing the share of interviewed individuals who stated that they could never imagine working abroad (‘‘share of stayers”).19 Following our argument, our main proposition is that an increase in migration prevalence in the region should increase the risk of trafﬁcking, while a higher share of stayers (lower migration pressure) should decrease trafﬁcking risks. On the supply side, one proxy for recruitment conditions in the region is the scale of illegal migration. The indicator used captures the regional proportion of migrants who work illegally abroad. Illegal migration is likely to foster the activities of people smugglers or agents procuring false travel documents, with which trafﬁckers can collaborate. And because shadow migration industries will tend to be larger the more people depart, the recruitment of illegal migrants may be particularly easy in emigration areas. We therefore expect the adverse effect of illegal migration to depend on migration prevalence rates, so that the interaction term of regional migration and illegality should be positively signed. The same intuition applies when using a dummy variable for illegal migration experience on the household level, which we also interact with migration prevalence. As a second demand-side proxy, we include a variable measuring the level of criminal activity (reported crimes per 10,000 inhabitants in a region) which we take from the Statistical Yearbooks of the respective countries and which we also interact with migration rates. We assume that trafﬁcking will be easier the more criminal a region, especially if this region has large emigration ﬂows. This would predict positive coefﬁcients of both the crime rate variable and its interaction with regional migration prevalence. Turning to the supply side of the human trafﬁcking market, our focus lays on indicators of desperation and negative selfselection into migration. Because the survey intentionally omitted questions on trafﬁcked family members, it is not possible to rely on education-based measures of selection. Likewise, we cannot construct a measure of network composition. However, the survey allows proxying the degree of desperation-led migration with the share of persons in the region who state that they would do ‘‘any kind of work” abroad.20 Economic deprivation can cause negative self-selection on its own, but this effect will be exacerbated in areas with high emigration and larger networks (McKenzie and Rapoport, forthcoming). To test this prior we interact the desperation proxy with migration prevalence and expect a positive coefﬁcient. Our intuition is that migrant networks will allow an even higher share of low-skilled and vulnerable workers to depart from deprived areas, resulting in more trafﬁcking. In a similar vein, we interact a dummy capturing a household’s ‘‘poor ﬁnancial status” with migration prevalence rates.21 We expect poorer workers, who might not have migrated in the absence of large migrant networks, to be more prone to fall victim to trafﬁckers. 17 An unavoidable bias in any database on human trafﬁcking or migrant exploitation is that it only provides information on victims with known whereabouts or those returned. Another problem might be that the survey did not put a focus on marginal social groups, such as Sinti and Roma, among whom trafﬁcking rates could be higher. 18 In light of the large economic and institutional differences between the countries under consideration, we cluster standard errors on the country level throughout the analysis. However, our main results are not affected when clustering on regions. 19 Since this variable could potentially be subject to reverse causality, the share is calculated excluding households with a trafﬁcking case in their close or distant family. Note that the share of people not wanting to leave is negatively correlated with regional migration prevalence rates, at a coefﬁcient of 0.30. 20 The question allowed for ten other answer categories. One can thus reasonably interpret the choice of ‘‘any kind of work” as a strong signal of desperation. To avoid endogeneity, this share is calculated excluding persons who live in migrant households or know victims of human trafﬁcking. 21 More precisely, households indicating that they ‘‘have to save money for the basic necessities (food, clothing, footwear)” are coded as poor. Unfortunately, this information is unavailable for households interviewed in Bulgaria and Romania.

180

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188

4.4. Other explanatory variables Two variables are used to assess the role of information asymmetries and access to information. First, we code a dummy for households that use the TV, as opposed to other media and social contacts, to get informed about social and political issues. TV use can be seen as a source of relatively objective and high-quality information and a potential channel for anti-trafﬁcking awareness campaigns. We therefore expect a negative coefﬁcient of the TV dummy. Second, we include a regional risk-awareness measure by taking the share of respondents who stated that they had heard of the phenomenon of human trafﬁcking before.22 Intuitively, risk awareness is likely to reduce the likelihood of trafﬁcking, which again predicts a negative coefﬁcient of this variable. Given the heterogeneity across regions and households, it is necessary to control for additional factors that might inﬂuence trafﬁcking risks.23 We proxy the level of socio-economic development with regional infant mortality rates (per 1000 live births) and the number of physicians (per 10,000 inhabitants). The remoteness of a region is captured by the share of people living in rural localities. On the household level, we use the number of children aged 16 or younger,24 a dummy for households living in rural areas and another dummy indicating whether households live in the district around their respective capital city. We also include age, gender, marital status, and educational level of the respondent, but only to control for systematic reporting biases along these dimensions. Descriptive statistics of all variables are presented in Table A1 in the appendix. 5. Results 5.1. Main results Table 1 displays our main estimation results for the whole sample.25 First, and most importantly, we ﬁnd migration prevalence to be the key predictor of human trafﬁcking (column 1). Compared to other explanatory variables, it has a large marginal effect.26 A one percentage point increase in the share of migrant households in a region raises the likelihood of having a trafﬁcked family member by 0.056 percentage points. This may appear small, but one has to keep in mind that trafﬁcking is a rare event. A one percentage point increase in migration prevalence effectively translates into a 5% increase in the predicted probability of human trafﬁcking on the household level. Or, when increasing the regional share of migrant households by one standard deviation (from 0.20 to 0.37), the risk of trafﬁcking more than doubles. This ﬁnding is conﬁrmed when we use our proxy for low migration pressure, the regional share of people who could never imagine working abroad (‘‘share of stayers”), see column 2. The incidence of trafﬁcking appears to be signiﬁcantly lower in regions where fewer people intend to leave. A further main result is on the role of risk perceptions and trafﬁcking. Households in regions with higher awareness of the phenomenon of human trafﬁcking are less likely to have a trafﬁcked family member. The same is true for households which use the TV as main source of information on social and political issues.27 These ﬁndings, although not fully robust, underline the potential beneﬁts of public awareness campaigns to counter human trafﬁcking. To some surprise, we ﬁnd that many regional control variables seem to play no role. Living in a region with a large share of rural population, a high rate of infant mortality, or low density of physicians does not increase the likelihood of trafﬁcking. If at all, there is weak evidence that households in regions with higher infant mortality are less likely to report a trafﬁcked family member. The second part of Table 1 shows our main results for the subsample of 1560 migrant households. The analysis for this subsample is of particular relevance, as it allows testing our hypothesis that individual trafﬁcking risks rise with regional migration rates. As can be seen, we ﬁnd that all of our key results hold in this migrant subsample. In particular, columns 3 and 4, show that the regional scale of emigration remains a key predictor of trafﬁcking. Migration prevalence has a positive and signiﬁcant coefﬁcient, indicating that trafﬁcking risks increase more than proportionally with emigration rates. We also ﬁnd that our proxy for low migration pressure (‘‘share of stayers”) keeps its signiﬁcantly negative coefﬁcient. 5.2. Results on the channels at work We have demonstrated that regional migration ﬂows are associated with a higher individual likelihood of being trafﬁcked. This section presents evidence on the channels at work. As our empirical strategy in this part of the analysis relies 22 We try to most closely resemble a migrant’s knowledge status before departure and reduce potential endogeneity by excluding respondents who knew a victim of human trafﬁcking or lived in a migrant household when calculating this share. 23 These variables were coded from the Statistical Yearbooks of each of the ﬁve countries. Given that our dependent variables are retrospective and in order to minimize potential endogeneity problems, all variables from the Statistical Yearbooks were coded for the year 2004. 24 Due to a modiﬁcation of the question, the number of children is deﬁned as 14 or younger in Romania. 25 All of our main ﬁndings hold when standard logit procedures are employed. For reasons of brevity, however, these results are not reported. 26 With the exception of dummy variables, all marginal effects are evaluated at the mean of the independent variables. 27 Arguably, TV use may also be a proxy for household wealth. Note, however, that this result remains unaffected when controlling for a household’s ﬁnancial status (column 5 of Table 2).

181

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188 Table 1 Baseline regressions. (1)

(2)

(3)

All households (n = 5513) Baseline

Migration prevalence

With regional share of stayers Baseline

Coef/se

Marginal effect

4.726*** (1.013)

0.056

TV use Number of children (0–16) Rural Capital district Share of rural population Infant mortality Density of physicians Crime rate Belarus Bulgaria Romania Ukraine Age (respondent) Male (respondent) No or primary education (respondent) Secondary education (respondent) Married (respondent) Observations

Coef/se

Marginal effect

**

Share of stayers Awareness of trafﬁcking

1.535* (0.889) 0.381** (0.157) 0.063 (0.071) 0.165 (0.206) 1.265** (0.616) 0.599 (1.189) 0.071*** (0.021) 0.010 (0.011) 0.000 (0.001) 0.208 (0.286) 0.767 (0.467) 0.768** (0.321) 0.436 (0.424) 0.009* (0.005) 0.185 (0.116) 0.352* (0.212) 0.079 (0.186) 0.289** (0.115) 5513

(4)

Migrant households (n = 1560)

0.018 0.005 0.001 0.002 0.024 0.007 0.001 0.000 0.000 0.003 0.011 0.008 0.006 0.000 0.002 0.004 0.001 0.003

1.391 (0.686) 1.857* (1.081) 0.372** (0.155) 0.071 (0.070) 0.204 (0.203) 1.394** (0.665) 1.869 (1.186) 0.040 (0.030) 0.005 (0.011) 0.001 (0.001) 1.194*** (0.273) 0.250 (0.375) 1.313*** (0.232) 0.474 (0.488) 0.009* (0.006) 0.192* (0.116) 0.334 (0.207) 0.030 (0.182) 0.302** (0.120) 5513

With regional share of stayers

Coef/se

Marginal effect

5.108*** (0.784)

0.158

0.006 0.017 0.005 0.001 0.003 0.029 0.024 0.001 0.000 0.000 0.023 0.011 0.003 0.012 0.000 0.002 0.004 0.000 0.004

2.674*** (0.921) 0.562** (0.237) 0.161 (0.114) 0.032 (0.264) 0.504 (0.651) 0.813 (1.056) 0.113*** (0.028) 0.004 (0.021) 0.002 (0.002) 0.937** (0.435) 1.465*** (0.467) 0.050 (0.377) 0.985*** (0.332) 0.007 (0.006) 0.153 (0.243) 0.126 (0.290) 0.006 (0.310) 0.138 (0.207) 1560

0.083 0.020 0.004 0.001 0.017 0.026 0.004 0.000 0.000 0.041 0.079 0.003 0.045 0.000 0.005 0.004 0.000 0.005

Coef/se

1.711* (0.947) 2.932*** (1.089) 0.522* (0.285) 0.181* (0.108) 0.014 (0.280) 0.699 (0.713) 2.557** (1.067) 0.083** (0.035) 0.014 (0.019) 0.000 (0.002) 0.223 (0.437) 0.402 (0.431) 0.518* (0.304) 0.112 (0.455) 0.008 (0.005) 0.182 (0.245) 0.067 (0.283) 0.066 (0.325) 0.186 (0.202)

Marginal effect

0.004 0.055 0.020 0.005 0.000 0.027 0.086 0.003 0.000 0.000 0.097 0.007 0.014 0.017 0.000 0.006 0.002 0.002 0.007

1560

Results of rare events logit estimation. The dependent variable is a dummy for human trafﬁcking incidence on the household level. Country-clustered standard errors in parentheses. *** Signiﬁcance at a 1% level. ** Signiﬁcance at a 5% level. * Signiﬁcance at a 10% level.

on interaction terms, we follow McKenzie and Rapoport (forthcoming) and use linear regression models (OLS) for estimation. The main purpose is to facilitate interpretation of the interaction term coefﬁcients, which is problematic in the non-linear models typically used for binary outcomes, such as logit or probit models (Ai and Norton, 2003). The left side of Table 2 presents results on demand-side effects. The interaction term between the share of illegal migration in the region and regional migration prevalence is positive and signiﬁcant (column 1). The same is true for the interaction with illegal migration experience in the household (column 2). Both results indicate that illegality has a particular risk-increasing effect in high-migration areas. In line with our arguments above, trafﬁckers seem to take advantage of larger shadow migration industries and illegal migration patterns. However, we do not ﬁnd any evidence that crime facilitates trafﬁcking in regions with high rates of out-migration. Column 3 shows that the interaction of reported crimes and migration rates in the region is insigniﬁcant and negatively signed. This non-ﬁnding might be due to measurement error and difﬁculties in comparing crime statistics across the ﬁve countries under investigation. Moreover, we lack alternative measures, such as the share of people convicted or the number of homicides in a region, which might provide a better indicator for relevant

182

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188

criminal activities. As to potential supply-side factors, the results are displayed in columns 4 and 5. We ﬁnd that our proxy for negative self-selection in the region (the share of people who would do any kind of job abroad), positively depends on migration prevalence. This is consistent with our hypothesis that migrant networks may foster the departure of more vulnerable types of migrants, particularly in deprived regions. Note, however, that the coefﬁcient of the interaction term is signiﬁcant only at the 10% level and not fully robust to speciﬁcation changes. On the household level, we ﬁnd no evidence for negative self-selection effects. The dummy on poor ﬁnancial status and the interaction is insigniﬁcant. It should be kept in mind, however, that the survey does not report any pre-migration assessments of living standards. But one should also recall that other proxies of regional development, such as remoteness, child mortality or the availability of physicians also turned out to be of no importance. Overall, we thus ﬁnd remarkably little evidence for supply-side factors. 5.3. Robustness checks To test the robustness of the results presented above, this section conducts a series of additional checks. In a ﬁrst step, we explicitly account for the fact that human trafﬁcking should only be observable for migrant households. To econometrically allow for this we estimate a probit sample selection model with a dummy for migrant households as dependent variable in the selection equation and a dummy for households with a trafﬁcked family member in the outcome equation. One way to identify this model is to ﬁnd a variable that is a signiﬁcant predictor in the selection equation, but excludable from the outcome equation. Our dataset offers only a limited choice of such instruments. Arguably, the most credible variable is a dummy indicating whether the household has an employed member at home. Our identifying assumption is that having a domestically employed household member relieves the household’s migration pressure, but does not directly affect the risk of being trafﬁcked. Having an employed household member turns out to be a highly signiﬁcant and negative predictor of having a family member abroad.28 The results of the probit selection model (columns 1 and 2 of Table 3) closely resemble those of the rare events logit estimation. In addition, our estimates do not appear to be biased due to sample selection. The Likelihood Ratio test of independent equations cannot reject the null hypothesis of no correlation between the error terms (p-value 0.56). We are aware that identiﬁcation may not be fully warranted with this approach. We implicitly assume negative self-selection into migration, as households with unemployed members are more likely to send a migrant. As discussed above, such negative self-selection effects might also affect trafﬁcking risks, especially in high-emigration regions, which raises concerns on whether our instrument is strictly excludable. However, we show that instrument choice does not affect the results in the Moldova case study (see Section 6). Moreover, an alternative form of identiﬁcation through functional form, thus relying on the non-linearity of the model. Re-estimating the regressions without an instrument yields similar results, with migration prevalence remaining a signiﬁcant and substantial determinant of trafﬁcking risks (not reported). We thus conclude that it is sufﬁcient to separate the sample selection model and focus on the outcome equation as done in Tables 1 and 2. Another issue of concern is the role of migrant destinations. It is possible that high migration ﬂows correlate with migration to destinations with poorer law enforcement or weaker anti-trafﬁcking laws, e.g. Russia or Turkey. While the survey does not contain information on migrant destinations, we take advantage of well-known migration patterns across countries to conduct a simple robustness check. Speciﬁcally, we divide the sample into sending countries with predominantly ‘‘Western” destinations (Romania and Bulgaria) and countries with more ‘‘Eastern”-oriented migration ﬂows (Belarus, Moldova, Ukraine). We ﬁnd that in both groups of countries regional migration prevalence remains a highly signiﬁcant predictor of human trafﬁcking risks (columns 3 and 4). In a next step, we restrict our sample to rural households. The rationale for this is that networks are typically thought to have stronger effects in rural areas. However, as shown in column 5, the results remain largely the same. We also ﬁnd little changes when explicitly controlling for proxies of regional institutions. More speciﬁcally, we use the share of respondents answering that trafﬁcking in the region mainly occurs (i) because of weak laws and poor law enforcement or (ii) because of corruption. As can be seen in column 6, we ﬁnd these two proxies to be insigniﬁcant. Finally, we analyze the differing effects of family networks (strong ties) versus community networks (weak ties). Many of the arguments on networks made so far would apply to weak-tie networks, but not to family networks, which are likely to be more protective when it comes to human trafﬁcking. The family network variable available to us simply captures the number of other members in the households that have migrated in the last three years. Family networks appear to play a protective role. Their effect, however, is far from being statistically signiﬁcant (column 7). The same is true if we include regional ﬁxed effects (not reported). However, a more in-depth analysis with better data on family migration patterns is needed to strengthen this conclusion. 6. Moldova case study To supplement the analysis and address some limitations of the ﬁve-country survey, this section presents a detailed case study of Moldova, the Eastern European country which is most affected by human trafﬁcking. A main difference to the ﬁve28 Most evidence on migration in the region indicates that households ﬁrst try to place a member in the domestic labour market and send a migrant only if this is unsuccessful. Generally, migration in the region is overwhelmingly motivated by the lack of income or job opportunities (88% of respondents in the survey state so). In Moldova, for example, only about a third of those migrating (36%) were active in the labour market before their departure (Omar Mahmoud and Trebesch, 2009).

183

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188 Table 2 Results on the channels at work. (1)

(2)

(3)

Demand-side effects Illegal migration (region) OLS coef/se Share of illegal migration in region Illegal migr. share* migration prevalence Illegal migration (household) Illegal hh migration* migration prevalence Crime rate in region* migration prevalence Desperation-led migr. (% willing to do any job abroad) Desperation-led migration* migration prevalence Poor ﬁnancial status of household Household poverty* migration prevalence Migration prevalence Awareness of trafﬁcking TV use Number of children (0–16) Rural Capital district Share of rural population Infant mortality Density of physicians Crime rate Country dummies Control for respondent’s characteristics Observations

(4)

(5)

Supply-side effects Illegal migration (household) OLS coef/se

Criminal activity (region) OLS coef/se

Desperation-led migration (region) OLS coef/se

Poverty (household) OLS coef/se

0.119 (0.077) 0.348* (0.195) 0.030 (0.023) 0.176*** (0.058) 0.001 (0.001) 0.208 (0.196)

1.223* (0.717) 0.008 (0.051) 0.008 (0.088) 0.099 (0.066) 0.118** (0.055) 0.024 (0.015) 0.009 (0.006) 0.004 (0.018) 0.018 (0.024) 0.035 (0.052) 0.003 (0.002) 0.000 (0.001) 0.000 (0.000) Yes Yes

0.144*** (0.049) 0.113** (0.058) 0.021 (0.015) 0.008 (0.006) 0.006 (0.017) 0.021 (0.024) 0.037 (0.050) 0.003 (0.002) 0.000 (0.001) 0.000 (0.000) Yes Yes

0.260** (0.105) 0.142* (0.073) 0.024* (0.014) 0.009 (0.008) 0.005 (0.012) 0.024 (0.027) 0.036 (0.053) 0.003 (0.002) 0.000 (0.001) 0.000 (0.000) Yes Yes

0.137*** (0.040) 0.119** (0.055) 0.024* (0.015) 0.009 (0.008) 0.003 (0.012) 0.025 (0.024) 0.033 (0.044) 0.003 (0.002) 0.000 (0.001) 0.000 (0.000) Yes Yes

0.259*** (0.032) 0.233*** (0.047) 0.039*** (0.011) 0.015* (0.008) 0.001 (0.020) 0.016 (0.027) 0.029 (0.024) 0.003** (0.002) 0.001* (0.001) 0.000 (0.000) Yes Yes

1560

1560

1560

1560

987

Results of ordinary least squares estimation. The dependent variable is a dummy for human trafﬁcking incidence on the household level. Country-clustered standard errors in parentheses. *** Signiﬁcance at a 1% level. ** Signiﬁcance at a 5% level. * Signiﬁcance at a 10% level.

country study presented above is the range of available data for Moldova, allowing us to conduct the analysis at less aggregated levels. First, we have access to a household survey with very detailed information on migrant characteristics and their migration patterns including the country of destination. Second, the Ministry of Economy and Trade provides a rich database on socio-economic characteristics of every community in Moldova. We can therefore investigate the drivers of human trafﬁcking on the individual, not household level and include locality-speciﬁc migration prevalence rates and control variables. Individual data come from a specialized household survey on migration and remittances, capturing 3916 households. The survey was commissioned by IOM and executed by the CBSAXA opinion research ﬁrm in summer 2008 (see Lücke et al. (2009) for further details). The impressive scale of emigration from Moldova is reﬂected by the fact that 1127 households in the survey (29%) had at least one member abroad. Although the survey instrument does not explicitly focus on trafﬁcking and exploitation, we can take advantage of one question on problems confronted abroad. This question was answered by almost all respondents and allows identifying the worst forms of migrant exploitation analogous to the deﬁnition used so far. Speciﬁcally, we code migrants as trafﬁcked if (i) they were locked by their employer abroad, (ii) their passport was taken away and/or (iii) they were subject to sexual exploitation. Thirty-one migrants in our sample meet these criteria. This share is notably smaller than in the ﬁve-country survey, a fact that can be attributed to survey design, in particular to the way trafﬁcking-related questions were asked, but also to the degree of retrospectiveness, as the question only referred to the survey year and the preceding one (2007 and 2008). To assure comparability, the Moldovan case study follows the ﬁve-country analysis as closely as possible. Some minor differences in the construction of variables are worth mentioning. Above all, instead of controlling for the characteristics of the respondent, we include migrants’ personal traits directly. Community migration prevalence is proxied by the share of people who are temporarily absent from a locality. The Moldovan Labour Force Survey shows that this is a meaningful

184

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188

Table 3 Robustness checks. Dependent variable

Migration prevalence

(3)

(4)

(5)

(6)

(7)

Heckman probit selection model

(1)

(2)

Bulgaria and Romania

Belarus, Moldova, Ukraine

With regional institutions

With family network

Trafﬁcking Coef/se

Migration Coef/se

Trafﬁcking Coef/se

Trafﬁcking Coef/se

Rural households only Trafﬁcking Coef/se

Trafﬁcking Coef/se

Trafﬁcking Coef/se

2.744*** (0.758)

1.847*** (0.213)

3.870*** (0.569)

4.842** (2.352)

5.788*** (0.466)

5.261*** (1.495) 0.239 (1.992) 0.126 (1.191)

5.278*** (0.786)

Weak laws and law enforcement Corruption Family network Awareness of trafﬁcking TV use Number of children (0–16) Rural Capital district Share of rural population Infant mortality Density of physicians Crime rate Country dummies Control for respondent’s characteristics Employed Rho Observations

1.180 (0.430) 0.231 (0.170) 0.089 (0.064) 0.014 (0.133) 0.193 (0.253) 0.375 (0.634) 0.048 (0.029) 0.000 (0.010) 0.001 (0.001) Yes Yes

***

0.081 (0.152) 0.055 (0.054) 0.061** (0.026) 0.099** (0.043) 0.021 (0.077) 0.081 (0.199) 0.010 (0.007) 0.000 (0.003) 0.001** (0.000) Yes Yes

***

0.409 (2.279) 0.087 (0.442) 0.139*** (0.023) 0.052 (0.345) 1.040** (0.524) 4.079* (2.315) 0.169* (0.096) 0.010 (0.045) 0.013*** (0.001) Yes Yes

3.302 (1.152) 0.864*** (0.175) 0.209* (0.111) 0.086 (0.368) 0.374 (0.779) 0.138 (0.570) 0.069 (0.105) 0.011 (0.025) 0.000 (0.000) Yes Yes

552

1008

0.931 (1.041) 1.754 (1.448) 0.166*** (0.032) 0.037 (0.028) 0.003 (0.004) Yes Yes

2.533 (1.302) 0.568** (0.237) 0.157 (0.117) 0.027 (0.274) 0.490 (0.645) 0.763 (1.014) 0.114*** (0.031) 0.003 (0.021) 0.002 (0.002) Yes Yes

0.039 (0.041) 2.647*** (0.860) 0.632*** (0.185) 0.136 (0.136) 0.040 (0.267) 0.577 (0.742) 0.759 (1.078) 0.113*** (0.027) 0.000 (0.021) 0.002 (0.002) Yes Yes

864

1560

1550

***

2.566 (0.768) 1.065*** (0.216) 0.168 (0.175)

*

0.115*** (0.043) 0.471 (0.864) 5513

Results of rare events logit and Heckman probit selection estimation. The dependent variable in columns 1 and 3–7 is a dummy for human trafﬁcking incidence on the household level. Country-clustered standard errors in parentheses. *** Signiﬁcance at 1% level. ** Signiﬁcance at 5% level. * Signiﬁcance at 10% level.

proxy as labour migrants represent more than 80% of those temporarily absent (Görlich et al., 2007). A major advantage of this variable is its representativity at the local level, as it is derived from administrative statistics and not survey-based. In addition to these changes, availability of doctors is now expressed as the distance to the next doctor and crime rates are replaced by perceived levels of personal security. Taken together, the baseline results for the Moldovan migrant sample are nearly identical to the ones for the ﬁve-country study (column 1 of Table 4). This strongly supports our argument and ﬁndings from above. Migration prevalence remains by far the most important determinant of trafﬁcking risks. This result holds when using ﬁxed effects for the 35 regions in Moldova, restricting the sample to rural areas (not reported), or including dummies for migrant destinations (column 2). Quite remarkably, migrants’ personal characteristics, such as age, sex and education, do not appear to matter. The same is true for destinations. Neither is migration to Russia or Italy, which together host about 80% of Moldovan migrants, associated with increased trafﬁcking risks. Again, we also investigate the differing role of family and community networks. The Moldovan survey data allow for a better measurement of family networks, as it contains information on the number of family members that migrated prior to the migrant him- or herself. As can be seen in column 3, we now ﬁnd the family network variable to be signiﬁcant and negative, indicating that previous migration experience of other household members can help lowering the risk of trafﬁcking and exploitation. This supports the view that our arguments may only apply to weak-tie, but not strong-tie networks. Robustness of the probit selection model is tested by using an alternative instrument. Speciﬁcally, we use occupational information on whether surveyed individuals are employed as construction workers in Moldova or were so prior to depar-

185

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188 Table 4 Results for Moldova case study. (1)

(2)

(3)

Main results

Migration prevalence Destination Russia Destination Italy Destination other CIS countries Other EU countries Family network Share of illegal migration (% in region) Illegal migr. share* migration prevalence (region) Illegal migration (migrant) Illegal migration (migrant)* migration prevalence Higher education (migrant)* migration prevalence Network composition (% highly-educated migr.) Network composition* migration prevalence (region) Age (migrant) Male (migrant) Secondary education (migrant) Higher education (migrant) Married (migrant) Number of children (0–16) Rural Capital district Share of rural population Infant mortality Distance to doctor Crime in locality Observations

Baseline

With destinations

With family network

Coef/se

Coef/se

Coef/se

7.519*** (2.507) 7.357*** (2.612) 7.633*** (2.575) 0.008 (0.753) 0.590 (1.047) 0.063 (0.977)

(4)

(5)

(6)

(7)

Demand-side effects

Supply-side effects

Illegal migration (regional share) OLS coef/se

Migrant education

Illegal migration (household) OLS coef/se

0.259** (0.123) 0.017 (0.045)

Network composition

OLS coef/se

OLS coef/se

0.199 (0.168)

0.662*** (0.137)

0.736 (0.712) 0.931* (0.542) 0.040 (0.060) 1.556*** (0.589)

0.003 (0.018) 0.460** (0.209) 0.115 (0.166) 0.121** (0.051) 1.818*** (0.528)

0.011 (0.025) 0.015 (0.025) 0.002 (0.026) 0.000 (0.000) 0.041 (0.390) 0.170 (0.394) 0.148 (0.410) 0.001 (0.006) 0.016 (0.587) 0.035 (0.597) 0.008 (0.593) 0.001 (0.008)

0.000 (0.000) 0.001 (0.006) 0.001 (0.010)

0.263 (0.716)

0.007 (0.010)

0.003 (0.009) 0.005 (0.011)

0.370 (0.495) 0.459 (0.480) 0.387 (0.474) 0.008 (0.008) 0.027 (0.247) 0.041 (0.241) 0.028 (0.247) 0.001 (0.004)

0.006 (0.008) 0.002 (0.004)

0.005 (0.008) 0.005 (0.007) 0.000 (0.004) 0.000 (0.003)

0.143 (0.553) 0.228 (1.016) 0.000 (0.014) 0.000 (0.001) 0.049 (0.054) 0.321 (0.435) 1679

0.003 (0.005) 0.005 (0.008) 0.004 (0.013) 0.001 (0.009) 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) 0.000 (0.001) 0.000 (0.001) 0.005 (0.006) 0.007 (0.008) 1679 1679

0.272 (0.683)

0.200 (0.550) 0.353 (1.078) 0.000 (0.015) 0.000 (0.001) 0.046 (0.053) 0.293 (0.440) 1679

0.255 (0.738)

0.006 (0.011)

0.175 (0.563) 0.006 (0.007) 0.204 (1.051) 0.007 (0.011) 0.000 (0.014) 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) 0.049 (0.054) 0.000 (0.001) 0.335 (0.432) 0.002 (0.007) 1679 1679

0.000 (0.000) 0.000 (0.006)

0.000 (0.000) 0.001 (0.006) 0.001 (0.010)

0.001 (0.009) 0.002 (0.019) 0.000 (0.000) 0.000 (0.000) 0.000 (0.001) 0.002 (0.007) 1679

Results of rare events logit and OLS estimation in a sample of 1679 migrants. The dependent variable indicates if a migrant has been trafﬁcked or not. Locality-clustered standard errors in parentheses. Signiﬁcance at a 1% level. ** Signiﬁcance at a 5% level. * Signiﬁcance at a 10% level.

***

ture. Given the building boom in Moldova’s main destination countries until 2008 and the resulting demand for cheap construction labour, construction workers have a much higher probability to emigrate (compare Omar Mahmoud and Trebesch, 2009). At the same time, there is no reason to expect construction workers to face a higher (or lower) risk of trafﬁcking, an intuition that is conﬁrmed by the data. Using this identiﬁcation approach, we ﬁnd the key result to hold in the sample selection model (not reported). As before, our estimates do not appear to be biased due to sample selection. The remaining columns of Table 4 presents evidence on why trafﬁcking risks increase with migration prevalence. To test for demand-side forces, we rely on indicators of illegal migration and crime. This time, however, we can use the proportion of migrants who crossed borders illegally as proxy for the extent of illegal migration. Illegal border crossing is arguably a more exogenous indicator than illegal work abroad and might also capture the size of the regional shadow migration industry more accurately. In line with our previous evidence, our results point to the existence of agglomeration effects, as the impact of illegality increases with emigration rates (columns 4 and 5). Again, the existence of criminal activities does not appear to increase the likelihood of trafﬁcking in high-migration areas (not reported). As regards factors on the supply side, the availability of migrants’ educational attainment allows for alternative identiﬁcation of negative self-selection and the possibility to analyze the effect of network composition. We do not ﬁnd that larger networks induce particularly vulnerable migrants to leave (column 6). However, network composition, measured as the proportion of migrants with higher education, seems to

186

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188

play a role. Migrants from regions with high fractions of low-skilled migrants face a particularly high risk when networks are large (column 7). This lends some credence to the hypothesis that network quality may decrease with network size and the degree of negative self-selection. However, more research with better data is needed to validate this point. Overall, our ﬁndings for Moldova are very similar to the ones of the ﬁve-country study and more in favour of demand-side channels.

7. Concluding remarks Our results for the Eastern European context have several implications. First, policymakers concerned about human trafﬁcking should acknowledge the important role of migration. We ﬁnd strong indication that trafﬁckers and exploitative employers are more prone to recruit victims from regions with large-scale emigration ﬂows. For this reason, we propose to target awareness campaigns and other anti-trafﬁcking policies to areas where migration ﬂows are high or on the rise, not to remote areas with no or little migration. Second, our indicative ﬁndings on the role of risk awareness and information use may provide some orientation for policy initiatives facing cost-beneﬁt trade-offs. In fact, the results give some reason for optimism that anti-trafﬁcking campaigns and measures to reduce information asymmetries in the migration process can indeed be a promising way to reduce the crime. Besides awareness campaigns, the ILO (2005) suggests to set up labour market information systems on jobs at home and abroad and the general use of model employment contracts. Third, our results document the close link between illegal migration patterns and trafﬁcking. Migrants who cross borders illegally and who depart from regions with more illegal emigrants face signiﬁcantly higher risks. One can argue that restrictive immigration policies exacerbate the problem, pushing would-be immigrants into illegality and fostering the emergence of shadow migration industries. However, a moderate increase of legal migration opportunities is unlikely to solve the problem given the magnitude of global migration pressure (Amin and Mattoo, 2006). In addition, trafﬁckers’ incentives would remain unchanged through such measures. Finally, we ﬁnd some indication that large community migration networks increase risks and may thus have ambiguous effects. This stands in some contrast to the existing migration literature, which generally advocates a risk-reducing role of networks. However, more research is needed to better understand the role of social networks for migration and trafﬁcking. As a general conclusion, it seems likely that the market of human trafﬁcking and the number of victims will continue to grow as long as migration pressure remains high. Acknowledgments We are highly indebted to George Gigauri from IOM for his support. We thank the Editor and an anonymous referee for providing comments that helped to considerably improve this paper. We also thank Satish Chand, Michael Clemens, Alexander Danzer, Dennis Görlich, Johannes Koettl, Ghazala Mansuri, Andreas Steinmayr, Rainer Thiele and participants of the 3rd IZA/World Bank Conference on Employment and Development in Rabat, the Annual Congress of the European Economic Association in Barcelona and the Annual Conference of the European Society for Population Economics in Sevilla for helpful comments and discussions. All remaining errors are our own.

Appendix Table A1

Table A1 Summary statistics for households with and without trafﬁcked members.

Regional characteristics Rural locality Capital district Share of rural population Infant mortality Density of physicians (per 10,000 inh.) Crime rate (crimes per 10,000 inh.) Migration prevalence Share of illegal migrants Share of stayers Awareness of trafﬁcking

All households (n = 5513)

Migrant households only (n = 1560)

No trafﬁcked member

With trafﬁcked member

No trafﬁcked member

With trafﬁcked member

Mean

Std. dev.

Mean

Std. dev.

Mean

Std. dev.

Mean

Std. dev.

0.48 0.18 0.38 10.93 35.40 160.92 0.28 0.28 0.51 0.79

0.50 0.38 0.22 3.75 16.24 107.93 0.17 0.17 0.12 0.14

0.56 0.21 0.51 11.08 28.84 115.89 0.42 0.31 0.48 0.75

0.50 0.41 0.26 2.57 12.51 77.72 0.19 0.17 0.11 0.18

0.55 0.14 0.48 11.87 29.41 122.28 0.37 0.30 0.49 0.79

0.50 0.34 0.24 3.72 15.43 80.46 0.18 0.17 0.12 0.15

0.61 0.10 0.58 11.39 27.09 104.41 0.46 0.35 0.46 0.70

0.49 0.30 0.23 2.56 12.25 65.21 0.20 0.17 0.11 0.19

187

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188 Table A1 (continued) All households (n = 5513)

Migrant households only (n = 1560)

No trafﬁcked member

With trafﬁcked member

No trafﬁcked member

With trafﬁcked member

Mean

Mean

Mean

Mean

Std. dev.

Std. dev.

Std. dev.

Std. dev.

Share willing to do any job abroad (regional) Rule of law (perception-based) Corruption (perception-based)

0.05 0.27 0.20

0.05 0.14 0.14

0.07 0.24 0.22

0.05 0.12 0.13

0.06 0.27 0.23

0.05 0.15 0.14

0.08 0.24 0.21

0.05 0.12 0.13

Household characteristics Family network (other migrants in hh) Illegal migration experience Number of children (0–16) TV use Poor ﬁnancial status Employed hh member (dummy)

0.25 0.09 0.46 0.83 0.41 0.46

0.92 0.29 0.75 0.38 0.49 0.50

0.86 0.40 0.63 0.78 0.48 0.49

1.85 0.49 0.84 0.42 0.50 0.50

0.90 0.34 0.55 0.85 0.45 0.41

1.58 0.47 0.83 0.35 0.50 0.49

1.30 0.60 0.76 0.79 0.48 0.50

2.14 0.49 0.91 0.41 0.50 0.50

43.86 0.45 0.29 0.43 0.28 0.60

17.71 0.50 0.45 0.50 0.45 0.49

41.50 0.47 0.27 0.46 0.27 0.67

16.04 0.50 0.45 0.50 0.45 0.47

42.32 0.44 0.29 0.41 0.30 0.63

17.14 0.50 0.46 0.49 0.46 0.48

39.99 0.49 0.31 0.46 0.24 0.67

16.38 0.50 0.46 0.50 0.43 0.47

Respondent’s characteristics Age Male No or primary education Secondary education Higher education Married

References Adams, Richard H., 2005. Remittances, poverty, and investment in Guatemala. In: Schiff, Maurice, Özden, Çag˘lar (Eds.), International Migration, Remittances, and the Brain Drain. World Bank/Palgrave Macmillan, Washington, DC. Ai, Chunrong, Norton, Edward C., 2003. Interaction term in logit and probit models. Economic Letters 80, 123–129. Akee, Randall, Basu, Arnab K., Bedi, Arjun S., Chau, Nancy, 2007. Determinants of Trafﬁcking in Women and Children: Cross-National Evidence, Theory and Policy Implications. Mimeo. Amin, Mohammad, Mattoo, Aaditya, 2006. Can Guest Worker Schemes Reduce Illegal Migration? Policy Research Working Paper Series 3828. The World Bank, Washington, DC. Aronowitz, Alexis A., 2001. Smuggling and trafﬁcking in human beings: the phenomenon, the markets that drive it and the organizations that promote it. European Journal on Criminal Policy and Research 9, 163–195. Bales, Kevin, 1999. Disposable People: New Slavery in the Global Economy. University of California Press, Berkeley. Basu, Arnab K., Chau, Nancy, 2003. Targeting child labor in debt bondage: evidence theory and policy prescriptions. World Bank Economic Review 17, 255–281. Bauer, Thomas, Zimmermann, Klaus, 1999. Assessment of possible migration pressure and its labour market impact following EU enlargement to Central and Eastern Europe. IZA Research Reports 3, Institute for the Study of Labor (IZA). Bauer, Thomas, Epstein, Gill, Gang, Ira N., 2007. The inﬂuence of stocks and ﬂows on migrants’ location choices. Research in Labor Economics 26, 199–229. Beine, Michel, Docquier, Frédéric, Özden, Çag˘lar, forthcoming. Diasporas. Journal of Development Economics. Bhabha, Jacqueline, 2005. Trafﬁcking, smuggling, and human rights. Migration Information Source: Special Issue on Migration and Human Rights. Blum, Jerome, 1957. The rise of serfdom in Eastern Europe. American Historical Review 62, 807–836. Borjas, George J., 1987. Self-selection and the earnings of immigrants. American Economic Review 77, 531–553. Bruckert, Christine, Parent, Colette, 2002. Trafﬁcking in Human Beings and Organized Crime: A Literature Review. Royal Canadian Mounted Police. Mimeo. Carrington, William J., Detragiache, Enrica, Vishwanath, Tara, 1996. Migration with endogenous moving costs. American Economic Review 86 (4), 909–930. Chand, Satish, 2008. Economics of Human Trafﬁcking: A Survey. The Australian National University. Mimeo. Chiquiar, Daniel, Hanson, Gordon H., 2005. International migration, self-selection, and the distribution of wages: evidence from Mexico and the United States. Journal of Political Economy 113, 239–281. Clemens, Michael, Montenegro, Claudio, Pritchett, Lant, 2008. The Place Premium: Wage Differences for Identical Workers across the US Border. Policy Research Working Paper Series 4671. The World Bank, Washington, DC. Collier, Paul, Hoefﬂer, Anke, 2004. Greed and grievance in civil war. Oxford Economic Papers 56, 563–595. Coomaraswamy, Radhika, 2001. United Nations Addendum Report to the Human Rights Commission regarding Mission to Bangladesh, Nepal and India on the issue of trafﬁcking of women and children. Cranford, Cynthia J., 2005. Networks of exploitation: immigrant labor and the restructuring of the Los Angeles janitorial industry. Social Problems 52, 379–397. Dessy, Stéphane, Pallage, Sylvain, 2006. Some surprising effects of better law enforcement against child trafﬁcking. Journal of African Development 8, 115–132. Domar, Evsey, 1970. The causes of slavery or serfdom: a hypothesis. Economic History Review 30, 18–32. Fearon, James D., Laitin, David D., 2003. Ethnicity, insurgency, and civil war. American Political Science Review 97, 75–90. Fiorentini, Gianluca, Peltzman, Sam, 1995. The Economics of Organised Crime. Cambridge University Press. Freeman, Scott, Grogger, Jeffrey, Sonstelie, Jon, 1996. The spatial concentration of crime. Journal of Urban Economics 40, 216–231. Freeman, Richard J., 1999. The Economics of Crime. In: Ashenfelter, O., Card, D. (Eds.), Handbook of Labor Economics, Volume 3C. Amsterdam, North Holland, pp. 3529–3571. Friebel, Guido, Guriev, Sergei, 2006. Smuggling humans: a theory of dept-ﬁnanced migration. Journal of the European Economic Association 4, 1085–1111. Gathmann, Christina, 2008. Effects of enforcement on illegal markets: evidence from migrant smuggling along the south-western border. Journal of Public Economics 92, 1926–1941. Glaeser, Edward L., Sacerdote, Bruce, 1999. Why is there more crime in cities? Journal of Political Economy 107, 225–229. Görlich, Dennis, Toman Omar Mahmoud, Christoph Trebesch (2007). Explaining Labour Market Inactivity in Migrant-Sending Families. Working Paper 1391, Kiel Institute for the World Economy. Görlich, Dennis, Trebesch, Christoph, 2008. Seasonal migration and networks – evidence on Moldova’s labour exodus. Review of World Economics 144, 107–133.

188

T. Omar Mahmoud, C. Trebesch / Journal of Comparative Economics 38 (2010) 173–188

Gozdziak, Elizbieta E., Collett, Elizabeth A., 2005. Research on human trafﬁcking in North America: a review of literature. International Migration 43, 99– 128. Harff, Barbara, 2003. No lessons learned from the holocaust: assessing risks of genocide and political mass murder since 1955. American Political Science Review 97, 57–74. Hausmann, Ricardo, Pritchett, Lant, Rodrik, Dani, 2005. Growth accelerations. Journal of Economic Growth 10, 303–329. ILO, 2005. A Global Alliance against Forced Labour. ILO, Geneva. IOM, 2006. Human Trafﬁcking Survey: Belarus, Bulgaria, Moldova, Romania, and Ukraine. December 2006. IOM Kyiv, Ukraine. King, Gary, Zeng, Langche, 2001. Logistic regression in rare events data. Political Analysis 9, 137–163. Kraler, Albert, Iglicka, Krystyna, 2002. Labour migration in Central and Eastern European countries. In: Laczko, F., Stacher, I., Klekowski von Koppenfels, A. (Eds.), New Challenges for Migration Policy in Central and Eastern Europe. Asser Press, The Hague, pp. 27–58. Lackso, Frank, Gramegna, Marco A., 2003. Developing better indicators of human trafﬁcking. Brown Journal of World Affairs 10, 179–194. Laczko, Frank, 2005. Data and research on human trafﬁcking. International Migration 43, 5–16. Langberg, Laura, 2005. A review of recent OAS research on human trafﬁcking in the Latin American and Caribbean region. International Migration 43, 129–139. Leblang, David, Satyanath, Shanker, 2008. Politically generated uncertainty and currency crises: theory, tests, and forecasts. Journal of International Money and Finance 27, 480–497. Leman, Johan, Janssens, Stefan, 2007. Travel agencies as a linking element for human smuggling and trafﬁcking from Eastern Europe. Migration Studies 166, 443–459. LexisNexis, 2008. LexisNexis Global Survey of Human Trafﬁcking Laws. July 2008. . Lücke, Matthias, Mahmoud, Toman Omar, Steinmayr, Andreas, 2009. Labor Migration and Remittances in Moldova: Is the Boom Over? IOM, Chisinau. Mansoor, Ali, Quillin, Brice, 2006. Migration and Remittances: Eastern Europe and the Former Soviet Union. The World Bank, Washington, DC. Martin, Nina, Morales, Sandra, Theodore, Nik, 2007. Migrant worker centers: contending with downgrading in the low-wage labor market. GeoJournal 68, 155–165. McKenzie, David, Rapoport, Hillel, forthcoming. Self-selection patterns in Mexico–US migration: the role of migration networks. Review of Economics and Statistics. Mora, Jorge, Taylor, J. Edward, 2005. Determinants of migration, destination, and sector choice: disentangling individual, household, and community effects. In: Özden, Çag˘lar, Schiff, Maurice (Eds.), International Migration, Remittances, and the Brain Drain. World Bank/Palgrave Macmillan, Washington, DC. Munshi, Kaivan, 2003. Networks in the modern economy: Mexican migrants in the US labor market. Quarterly Journal of Economics 118, 549–599. Naim, Moses, 2005. Illicit: How Smugglers, Trafﬁckers and Copycats Are Hijacking the Global Economy. Doubleday. Obuah, Emmanuel, 2006. Combating global trafﬁcking in persons: the role of the United States post-september 2001. International Politics 43, 241–265. Omar Mahmoud, Toman, Trebesch, Christoph, 2009. Emigration and Skills Shortages in Moldova. Background Paper for Country Economic Memorandum. The World Bank, Washington, DC. Orrenius, Pia, 1999. The Role of Family Networks, Coyote Prices and the Rural Economy in Migration from Western Mexico: 1965–1994. Working Paper 9910, Federal Reserve Bank of Dallas. Orrenius, Pia, Zavodny, Madeline, 2005. Self-selection among undocumented immigrants from Mexico. Journal of Development Economics 78, 215–240. Piper, Nicola, 2005. A problem by a different name? A review of the research on trafﬁcking in South-East Asia and Oceania. In: Laczko, F., Gozdziak, E. (Eds.), Data and Research on Human Trafﬁcking: a Global Survey. IOM, Geneva, pp. 202–233. Ruggiero, Vincenzo, 1997. Trafﬁcking in human beings: slaves in contemporary Europe. International Journal of the Sociology of Law 25, 231–244. Salt, John, Stein, Jeremy, 1997. Migration as business: the case of trafﬁcking. International Migration 35, 467–494. Schloenhardt, Andreas, 1999. Organized crime and the business of migrant trafﬁcking. Crime, Law and Social Change 32, 203–233. Stalker, Peter, 2000. Workers without Frontiers: The Impact of Globalisation on International Migration. ILO, Geneva. Surtees, Rebecca, 2005. Second Annual Report on Victims of Trafﬁcking in South-Eastern Europe. IOM, Geneva. Tamura, Yuji 2007. Migrant Smuggling. Warwick Economics Research Paper Series 791. University of Warwick, Department of Economics. Tyldum, Guri, Brunovskis, Anette, 2005. Describing the unobserved: methodological challenges in empirical studies on human trafﬁcking. International Migration 43, 17–34. Tyuryukanova, Elena, 2005. Forced Labour in the Russian Federation Today. ILO, Geneva. UN, 2000a. Protocol to Prevent, Suppress and Punish Trafﬁcking in Persons, Especially Women and Children, Supplementing the United Nations Convention Against Transnational Organized Crime. United Nations, New York. UN, 2000b. Protocol against the Smuggling of Migrants by Land, Sea and Air, Supplementing the United Nations Convention against Transnational Organized Crime. United Nations, New York. UNODC, 2006. Trafﬁcking in Persons: Global Patterns. United Nations Ofﬁce on Drugs and Crime. US State Department, 2008. Trafﬁcking in Persons Report. Ofﬁce to Monitor and Combat Trafﬁcking in Persons. Väyrynen, Raimo, 2005. Illegal immigration, human trafﬁcking and organized crime. In: Borjas, George, Crisp, Jeff (Eds.), Poverty, International Migration and Asylum. Palgrave Macmillan, New York, pp. 143–170. Winters, Paul, de Janvry, Alain, Sadoulet, Elisabeth, 2001. Family and community networks in Mexico–US migration. Journal of Human Resources 36, 159– 184. World Bank, 2005. Global Economic Prospects 2006: Economic Implications of Remittances and Migration. Washington, DC. Zenou, Yves, 2003. The spatial aspects of crime. Journal of the European Economic Association 1, 459–467. Zhang, Sheldon, Chin, Ko-Lin, 2002. Enter the dragon: inside Chinese human smuggling organizations. Criminology 40 (4), 737–767.

The economics of human trafficking and labour ... - WordPress.com

human trafficking pathfinder

Human-Trafficking-eBook-v6web.pdf

Combating-Human-Trafficking-A-Multidisciplinary-Approach.pdf ...

D1) HUMAN TRAFFICKING pw.pdf

$pdf-1594\prostitution-and-human-trafficking-by-andrea-di-nicola ...$

pdf-1594\prostitution-and-human-trafficking-by-andrea-di-nicola ...

The Transformation of the Value (and Respective Price) of Labour ...

India Labour and Employment Report 2014 - Institute for Human ...

Human Trafficking: Modern-Day Slavery in America | WGBH News.pdf ...

National Human Trafficking Resource Center Staff ... -

Labour mobility and the redistributive effects of trade integration.pdf

Trafficking and Forced Prostitution of Palestinian Women and Girls ...

Human Trafficking Charities lying to get government Grants.pdf ...

National Human Trafficking Resource Center Staff ... -

Trafficking and Forced Prostitution of Palestinian Women and Girls ...